The first development stage is complete! (Results)
Omega Server
Spread the love

PANN

Discussing the results of the first development stage of a new generation neuroserver.

The duration of the first stage - November 1, 2020 - May 1, 2021.

This report provides textual and video information.

The main result of the first stage – the functioning Omega-Server platform is open for interested individuals and enterprises, where you can experiment with PANN, experience its unique features and achieve your goals with PANN.

Video explaining how to work with the alpha version of the platform: link

To download the distribution kit, please register on our website here: Omega-Server Platform

Experts can download information about PANN testing using the IRIS and MNIST databases from GitHub.

Neural network designers usually use different databases to test their networks.

Geoffrey Hinton has figuratively described MNIST as the “fruit fly of machine learning,” which means it allows machine learning researchers to study their algorithms in a controlled laboratory setting, as biologists often study fruit flies.

IRIS: https://github.com/Omega-Server/p-net-comparison-iris

MNIST: https://github.com/Omega-Server/p-net-comparison


For professionals, we would recommend a fundamental work on neural networks: Deep Learning. By Ian Goodfellow, Yoshua Bengio, and Aaron Courville.

Other notable experts in the field who have shared this recommendation are:
• Elon Musk, OpenAI Co-Chair; co-founder and CEO of Tesla and SpaceX
• Jeffrey Hinton FRS, Emeritus Professor at the University of Toronto; Distinguished Research Fellow at Google
• Yann LeCun, Director of AI Research, Facebook; Silver Professor of Computer Science, Data Science, and Neuroscience, New York University.

You can study all the materials in this report or start with a topic that interests you.

Don't miss your unique chance to be at the forefront of this new key and versatile technology!

If you are still not sure how widely neural networks are already in use, check out the list below ...

In the commercial sector of the economy, AI technologies are widely used to solve the following problems:
– recognition and translation of multilingual speech streams in the presence of noise;
– text recognition, recognition of emotions in the text, prediction of the continuation of a phrase, translation of the text;
– creation of original texts in a living language from a large amount of data;
– cryptography (decryption of texts), decoding of genomes;
– merging 2D images into 3D (e.g., cartography, tomography, radiology);
– face recognition, recognition of emotions from a photograph, determination of the authenticity of a photograph, recognition of handwritten texts, recognition of areal images;
– forecasting the strength of buildings and structures;
– intelligent training systems;
– financial planning, sales, stock market games, management of securities portfolios, loan assessment (assessment of the applicant's creditability and the size of the loan);
– logistics (ordering and managing the movement of goods and components);
– analysis of complex data in medical systems, diagnostics and prescription of treatment, selection of drugs, prediction of mental states;
– search for malware;
– search for cyber vulnerabilities;
– games, including card games, with the possibility of bluffing;
– diagnostics of technical systems;
– legal advice and criminal proceedings, including the forecast of a court decision of a particular judge;
– autonomous and automated navigation (vehicle traffic control) in 3 physical environments;
– flexible traffic control (traffic light control);
– proof of theorems, formation of hypotheses, formation of knowledge bases for expert systems;
– synthesis of complex objects: synthesis of new drugs, synthesis of complex organic compounds with desired properties, synthesis of genomes for new organisms.
The total number of software products for each position is very large.
For more, please, see: О возможных применениях искусственного интеллекта в КСА войск

PANN properties and differentiation – a conversation with Vladimir Proseanic, CEO and Chairman of the Board of Directors of Omega Server Inc.

   

PANN unique features

 

For non-professionals
Article for review: https://omega-server.ai/en/neuron-new-type/

For professionals - a conversation with Boris Zlotin, Chief Scientist of Omega Server Inc.


Also for professionals: the article A new approach to creating neural networks link, PDF

Demonstration of the Omega-Server platform.

The completion of the first stage has created the basis for further increasing the value of the company:

  1. The Omega-Server platform and the formed working team allow us to create various AI applications.
  2. We have developed marketing tools - PANN demo and free access to independent testing by potential clients who can also experiment with PANN as applied to their objectives. This allows potential clients to independently verify the declared characteristics of the PANN technology.
  3. A basis has been created for the development of breakthrough technologies, for example, for the creation of brand new "smart databases" with great potential for applications and sales.

As a result, the company’s share price increased to $5.00 per share.

Examples of PANN application

Any PANN application is beneficial for the following reasons::

  1. A small number of training examples is needed: instead of thousands or tens of thousands of samples - tens or hundreds. There is a time saving on the formation of a training set - days, weeks and months.

  2. The training time is hundreds of times less. The larger the training dataset, the greater the difference in training time for our network from the classical one. Saving developer time - days, weeks and months.

  3. Our networks allow you to work effectively on computers with an order of magnitude less computing power, for example: instead of a workstation - a desktop can be used, instead of a server - a workstation. Substantial savings in hardware and energy costs.

  4. PANN can be permanently additionally trained to new data only, improved, customized, quickly adapted to customer requirements, which allows to achieve a serious marketing advantage. PANN is a transparent system that is easy to use, and any programmer can do it.

Examples:

PANN Industrial Applications
This application can be used for various industrial purposes, for example:
• Quality control of manufacturing operations, such as:
- Stamping
- Machine processing
- Welding
- Painting or enameling
- Assembly
- Packaging, etc.
• Control of many elements of assembly and welding.
• Control over the fulfillment of work assignments by personnel.
• Control of personnel performance with various objects.
• Risk management and prevention for organizations, buildings and structures.

 


Text recognition - chatbot
Chatbot: The application can be used for correspondence between the user and the network on various topics. It can also be used to organize dialogues between AI and clients, for example:
• Technical support of sales, including hardware, software, and other products.
• Providing service responses – for government agencies, phone providers, banking services.
• Providing stock exchange news, trading tips.
• Much more...

 

Text recognition – article titles (headlines) and their feedback
Analysis of article titles (headlines):
• Allows you to form effective headers, providing positive feedback.
• Analysis can be applied to the entire text.
• Identification of the necessary information, for example, for the compilation of specialized databases in the areas of interest.
• Indexing information base for fast and correct information search.

 

Face recognition
Multiple applications:
• Identification of criminals and mentally unstable persons who pose a danger to society.
• Search for the missing.
• Analysis of information from video cameras – though video cameras are widely available, there is no video analysis of streaming information.

 

Recognition of human emotions
Hand gesture recognition
Allows one to enter information by "drawing in space" in front of the video camera. This can be useful when entering information into a computer and for controlling any equipment using hand gestures.

 

Recognition of finger movements
Human posture recognition
Human postures allow us to analyze the physical or emotional state of a person. For example, in sports - to determine the correctness of the exercise. This can be very helpful for a virtual trainer.

 

Recognition of various objects
Unlimited applications: fruits, vegetables, cars, boxes, jewelry, jackets, shirts, dresses, etc.

 

Speech recognition
Applications - any area where voice control of hardware or software is needed:
- Entering information.
- Computer, vehicle, other equipment control.
- House control, including lights, appliances, doors, etc.

 

Voice command recognition

These capabilities of our network allow everyone to create their own software to complete any task. The most promising is the creation of software that integrates different applications.

Application of matrix algebra
In the very first phase of development, we started working ahead of schedule on the application of matrix algebra to our network. Matrix algebra gives us several significant advantages over existing networks:
1. The training speed increases many times over, since our networks do not require complex mathematical calculations.
2. The use of GPUs (graphic cards or graphics accelerators that allow you to work in parallel with many threads of computation) will speed up computations by hundreds and thousands of times.

 

The training speed of a neural network, achieved by using matrix algebra on a GPU server, will allow you to work with huge databases that only supercomputers can handle today.

 

Prospects: science fiction becomes reality; a civilizational turnaround is on the way.

Immediate prospects -
II stage of development.

  

Introducing the PANN Development Team (New Key Players)

Options for cooperation:

- if you are an entrepreneur and envision how to benefit from PANN technology; if you have good neural network and software specialists, we can establish mutually beneficial cooperation. To start the communication, please, fill out the form:

QUESTIONNAIRE FOR AN ENTREPRENEUR

    - if you are a scientist interested in artificial neural networks, you can be at the very forefront of the PANN development.

    To start the communication, please, fill out the form:

    QUESTIONNAIRE FOR A SCIENTIST

      - if you are an investor, please, review our investment offer.